Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
So Elon Musk thinks putting phones with ChatGPT enabled in a Faraday Cage (I had no idea what that was) will stop the AI from working?
IMG_0107.jpeg
 
  • Like
Reactions: IllegitimateValor
So Elon Musk thinks putting phones with ChatGPT enabled in a Faraday Cage (I had no idea what that was) will stop the AI from working?
View attachment 2387308
Faraday Cages are about blocking RF (radio frequencies). It’s more about blocking cellular communications back to OpenAI’s servers and preventing inadvertent data exfiltration. I know it’s fun to hate on Elon Musk, but there is SOME logic to his complaint here.
 
It’s worth putting Musk’s comments in the context of the major falling out over the direction of OpenAI which he was one of the co-founders, along with a few others including Sam Altmanwnd Greg Brockman. Originally it was a non-profit organisation to further AI and technically still is but they set up a for profit subsidiar, OpenAI Global, which is the part Microsoft invested in to enhance Bing and develope Copilot. Musk has accused OpenAI of moving away from the reason it was set up and is currently embroiled in a lawsuit with OpenAI. In that context his comments aren’t unexpected.
 
Faraday Cages are about blocking RF (radio frequencies). It’s more about blocking cellular communications back to OpenAI’s servers and preventing inadvertent data exfiltration. I know it’s fun to hate on Elon Musk, but there is SOME logic to his complaint here.
Thought it was a Friday Cage when I first read it.
 
Last edited:
  • Like
Reactions: kc9hzn
Faraday Cages are about blocking RF (radio frequencies). It’s more about blocking cellular communications back to OpenAI’s servers and preventing inadvertent data exfiltration. I know it’s fun to hate on Elon Musk, but there is SOME logic to his complaint here.
Nope. No logic. Device goes into cage -> device realizes there's no net connection -> device accumulates local queue of transactions -> device leaves cage -> device goes through queue. In the end, nothing is gianed.

Note here that Elon's concern is the security of proprietary data in his own companies. His concern is markedly NOT for the private information of users, or protecting them from misinformation or exploitation.

Apple gives a tin **** about the privacy of user data. Sure it's made of tin, but: Other companies, including those run by Elon, are _user_privacy_hostile_. If he could legally demand that you submit your own personal phone for deep inspection by his security personnel every time you show up to work, he would do so.
 
  • Like
Reactions: IllegitimateValor
it's already been reported he left because he believe OpenAI wouldn't get to AGI without major hundred billion dollar investments.
Technically it's been reported that:
* He left OpenAI because on conflict of interest for poaching
* He left OpenAI because they refused his merger with Tesla
* He left OpenAI because they wanted to become for-profit
* He left OpenAI because they wanted to go closed source

His statement in an email one year after he left the board probably isn't a reliable indicator of the reason he left. I assume you're referring to his statement:

"Even raising several hundred million won’t be enough. This needs billions per year immediately or forget it"

Honestly, if the reason he actually left was they wouldn't be able to create a theoretical Artificial General Intelligence, when the applications of OpenAI are already massive - then that's the dumbest thing I've ever heard.

"Oh you've created a service that every business wants to utilise, but you haven't made a virtual god in a few years and it'll be rather expensive to do so? Well I'm leaving to work on my self-driving cars that will be ready... uh... next year, which is a statement I'll repeat 10 years in a row"
 
He has a point. How can we verify that our data won’t be made accessible to OpenAI? I trust Apple. I don’t trust OpenAI.
Can you therefore trust Apple to check out OpenAI and make sure they aren't doing anything they shouldn't?
I don't see Musk attacking Android for Gemini which confirms for me that this is just his beef with OpenAI not giving him what he thinks he owns - the secret sauce of ChatGPT which his grok can not match. ChatGPT is in the lead and he knows it! He needs that for his self driving cars and other projects.
Keeps telling us it's better but...
 
  • Like
Reactions: IllegitimateValor
Typically reactionary fear-mongering, which should be expected from someone with Musk’s ideology. In reality he’s just salty about being booted from OpenAI and having a much crappier version with his own company. Meanwhile X continues to burn, overwhelmed with bots, porn, white supremacists, and sometimes all three in one!
 
I'd rather porn be banned on X.
agreed partly- Some of the only-fans-funded repliers are the funniest there; what needs to be eliminated are the pr0n bots- all the “udes in profile” ghost account replies under everything. And ever since he let people pay to move to the top rank of replies it’s ruined the Twitter experience. Let the likes and replies decide- not someone’s willingness to shell out money
 
Nope. No logic. Device goes into cage -> device realizes there's no net connection -> device accumulates local queue of transactions -> device leaves cage -> device goes through queue. In the end, nothing is gianed.

Well, never mind the fact that a phone in a Faraday Cage can’t actively record audio from outside of the cage or accept touch events while in the cage. Kinda hard to use a phone when it’s in a 360° metal vault, after all. Device does accumulate a local queue of transactions, but those transactions, crucially, don’t contain information from the user’s work environment (on account of the device being locked up in a metal box and not at the user’s desk).
 
Because Apple would be liable for making false claims and sued to the ends of the earth. Does Elon realize that any device that is bought into his facilities can run ChatGPT?
Wrong. First, you would need the technical know-how to detect and prove that Apple's 'promise' was broken. Then, you would have to prove that you, personally, were harmed. That harm would have to be substantial and specific. Without all that, you wouldn't even have standing to sue Apple for anything. So, no, people 'suing Apple for false claims' is no assurance of security. Only Apple's own competence, and desire to protect its reputation, can provide that.
 
Technically it's been reported that:
* He left OpenAI because on conflict of interest for poaching
* He left OpenAI because they refused his merger with Tesla
* He left OpenAI because they wanted to become for-profit
* He left OpenAI because they wanted to go closed source
well no, he did some of those things but that's not why he left. he left because Elon said there needs to be a relevant competitor (which will require large amounts of capital as he stated in previous email) and he said he was going to do it himself.

and Elon did want to go for-profit and agreed that Open in OpenAI did not explicitly mean open sourcing (and was ok with parts of it being closed source) but rather Open means everyone should benefit from the work OpenAI is doing.
 
  • Like
Reactions: Sybil24
I like Elon, what he's done at Tesla and all his companies is amazing, including X. People criticizing him here are foolish, blinded by the current political echo chamber.
That being said, this seems to be an overblown take and I disagree with him here. Some mistrust of OpenAI is warranted, but this seems biased. Either he is just reacting without thinking, or trying to create controversy for leverage for Apple integrate Grok down the line. Honestly I'm fine with Grok, although not his comments, more choice of AI models the better. I trust Apple to do the right thing, even with OpenAI.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.